CHILDS GLACIER 411 Altogether the 1911 conditions indicated a rapid return to the normal state which prevailed before the advance of 1910. It would have been scientifically interesting, though unfortunate for the railway, to have seen the advance continue during the low water stage from October, 1910, to June, 1911, with a gradual forcing of the river eastward, the cutting back of the eastern river bank in the unconsolidated gravels, the interference with stream flow and iceberg discharge from Miles Glacier, and many accessory phenomena. Fortunately the glacier advance came at the time of high water, and was nearly over by the beginning of the low water stage in October. The advance would have easily reached the railway if it had occurred in early spring, late fall, or winter. It seems probable, however, that the river can always be depended upon to protect the bridge and railway, if future advances come at high water, when undercutting of the glacier seems competent to cause enough iceberg discharge to keep back the advancing ice front; unless, of course, there occurs a period of prolonged or far more active advance. It is unfortunate that there is no similar protective relationship of the river to the Allen, Grinnell, Heney, and Kennicott Glaciers, whose outer portions are also traversed by this railway and which may sometime have advances similar to that of Childs Glacier in 1910. Cause of the 1910 Advance. The advance of Childs Glacier in 1910 is of interest in connection with the spasmodic advances or glacier floods of the Yakutat Bay glaciers since 1899. The recent history of the ice tongue may be summarized as follows: Omitting the imperfectly-known history of Childs Glacier at the time of the visits by the Russians, we know that it was engaged in what seems to be normal flowage from 1884 to 1909, interrupted by a slight increase in rate of movement in 1905-6. In 1909 this normal movement was interrupted by a marked increase in rate of flow but this did not assume great proportions until the summer of 1910. During t.Tn'q summer the advance was phenomenal, being similar to what we infer to have taken place in the Yakutat Bay. glaciers during' their brief spasmodic advances, though with a longer total period of advance, in this respect resembling the conditions in the Vernagt-Ferner of the Tyrol. The Childs Glacier began its unusual activity in 1909, was many times more active in 1910, slowed down by October, 1910, but continued a slight advance during the winter of 1910-11. In June, 1911, the conditions seemed to be nearly normal again,'though there was still some advance.1 This shows clearly that a great wave of motion surged down through Childs Glacier, beginning to make itself felt in 1909, causing greatest activity in 1910, and becoming greatly reduced in about a year from the time it commenced. Mr. E. C. Hawkins was told by a reliable prospector, who witnessed the happening, that during an unusually warm, rainy period in January or February, 1910, a detached hanging glacier fell from the walls of the Childs Glacier valley about 10 miles west of Copper River. A stupendous quantity of ice was avalanched out upon the glacier surface. This of itself could not have caused the advance of 1910, which indeed had already commenced before the avalanche. It illustrates, in a small way, however, what may take place on a much larger scale when a severe earthquake shakes a mountain region like that near Childs Glacier, which has many detached glaciers hanging on steepened slopes and much snow in unstable equilibrium. i Caleb Corser of Cordova observed in 1012 that Childs Glacier had advanced 60 feet from July 15 to September SO. It is presumed that this was measured at the northern margin (see table, p. 405).